Computer analysis of eye movements during reading

Publisher Website
Google Scholar
Abstract
This paper describes the application of a computer‐based analysis system to the evaluation of electrooculographically recorded eye movements during reading. Through the technique described, a more precise description of the activity of the eyes can be obtained for analysis. As an illustration of the use of this technique, a study is described wherein a number of hypotheses associated with fixation duration of specified eye movements during reading were tested. The results demonstrate that in relatively skilled readers fixation duration for the fixation pause prior to, during, and immediately following a regressive eye movement are significantly shorter than those associated with “normal” reading patterns. The first fixation pause following the return of the eye to a new line is significantly longer than those associated with “normal” reading, while duration of the last fixation pause on a line is more variable than the other fixations. The study identifies five eye movement patterns which can be discriminated both qualitatively and quantitatively from normal forward‐going reading patterns. Controversy exists over the applicability of analyses of eye movement to the diagnosis and remediation of reading. Some authorities believe this approach to be of little or no use (Zintz, 1966) while others express optimism about the potential usefulness of such analyses (Ahrendt, 1971). We believe with the latter author that more refined description and micro‐analysis of eye movements are essential to an understanding of the more complex interrelationships involved in the complex task of reading. Clearly, description and analysis are dependent upon accurate measurement if they are to be of practical use. For this reason, we further contend that one of the bases for the lack of general enthusiasm for studying eye movements in reading research is attributable to the relatively crude nature of the analysis typically performed. The use of computers in the collection and analysis of eye movement data, as herein described, overcomes a number of these obstacles: first, it takes much of the drudgery out of the analytical process, and second, it allows a more precise analysis of the visual search process which occurs during reading than has heretofore been possible. This paper describes the development and nature of the computerized analysis system, and provides an illustrative example of the use of such a system in eye movement research.

This publication has 1 reference indexed in Scilit: